Objectives: We aimed to explore the association between vitamin D levels and the severity, mortality and microbiological etiology of community-acquired pneumonia. Methods: Vitamin D levels (both, the reservoir form 25-OH and the activated form 1,25-OH2) of 300 randomly selected patients with community-acquired pneumonia due to pre-specified pathogens included in the German competence network (CAPNETZ) study were measured. Prior to statistical analysis, values of 25-OH and 1,25-OH2 were power-transformed to achieve parametric distribution. All further analyses were performed with seasonally and age adjusted values. Results: There was only a modest (Spearman Coefficient 0.38) positive correlation between 25-OH and 1,25-OH2. For 1,25-OH2 but not 25-OH, the general linear model revealed a significant inverse correlation between serum concentration and CURB score (p = 0.011). Liver and respiratory co-morbidity were associated with significantly lower 25-OH values and renal co-morbidity with significantly lower 1,25-OH2 values. No significant differences of 1,25-OH2 or 25-OH between different pathogens (influenza virus, Legionella spp., Streptococcus pneumoniae) were detected. Conclusion: For 1,25-OH2, we found a significant and independent (controlled for age, season and pathogen) negative correlation to pneumonia severity. Therefore, supplementation of non-activated vitamin D to protect from pneumonia may be non-sufficient in patients that have a decreased capacity to hydroxylate 25-OH to 1,25-OH2.